Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: error message after using -reoprob-


From   "Gindo Tampubolon" <Gindo.Tampubolon@manchester.ac.uk>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: error message after using -reoprob-
Date   Fri, 25 Apr 2008 10:58:02 +0100

Dear Li,

Sorry to put cold water in.
I do not think dynamic random effect ordered probit is a matter of adding lag of dependent variable into independent vars set and estimate standard random effect ordered probit. At least that's one reading of Heckman 1981 'The incidental parameters problem and the problem of initial conditions in estimating a discrete time-discrete data stochastic process.' In Manski and McFadden "Structural Analysis with Discrete Data with Econometrics Applications." [By the way you can google and get the full book.]


A good while ago, please check the Stata archive, Wiji Arulampalam nee Narendranathan started with similar question, I think, regarding dynamic random effect probit. Note: not even ordered probit.
To cut the long story short, to deal with incidental parameters problem, we [Stata users] now have the wonderful though painfully slow -redpace- and -redprob- thanks to Mark Stewart.

Maybe somebody somewhere have programmed/worked out the estimation and inference of dynamic random effect ordered probit AND dealt with incidental parameters problem; and by doing so point out that I've been confused all along. I'd appreciate that.

Cheers

Gindo
University of Manchester

----------------------------------------------------
Date: Thu, 24 Apr 2008 23:13:27 -0500
From: Richard Williams <Richard.A.Williams.5@ND.edu>
Subject: Re: st: error message after using -reoprob-

At 08:53 PM 4/24/2008, J. Li wrote:
>Hi all,
>
>I am using -reoprob- to estimate a dynamic ordered probit panel data
>model with random effects. The dependent variable is self-assessed
>health status with 5 ranks and only the first lag is included in the
>regressor list. I got some strange error message for some of my
>regressions. The command I used is as following:
>
>xi: reoprob $yvar hlthclag1 hlthclag2 hlthclag3 hlthclag5 agec genderc
>i.area fsize birthage lnhincome schoolm2 schoolm3 schoolm4 schoolf2
>schoolf3 schoolf4 pmkfe pmknm mlnhincome mschoolm mschoolf if
>qpropeducu==1, i(persruk)
>
>
>The error message showed up when fitting the constant-only model as
>following:
>
>__00000W not found
>
>
>What might be the problem here? Thanks very much!

There is a temporary variable it cannot find for some reason.  Before
running reoprob, you can give the command -set trace on- and then
look to see what is happening when the program dies.  The problem may
or may not be obvious.

I would also use the actual y variable name rather than $yvar, and I
would compute the area dummies beforehand rather than use xi.  It
probably won't matter, but then again the simpler your syntax perhaps
the less likely it is that reoprob will make a mistake.

You might also try using regoprob, available from SSC.  Same syntax
as what you have, except add the -pl- option at the end of your command.


- -------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW:    http://www.nd.edu/~rwilliam

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

------------------------------

End of statalist-digest V4 #3042
********************************

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index